behnamoh

17 ms·

361.
by behnamoh 10mo ago
This is what happens when a system is designed by multiple people and companies over a long period of time. An amalgam of ideas which are there just because. There's no reason Linux should be like this. e.g., see https://gob
362.
by behnamoh 10mo ago
the same CNN is sleeping on the Iran developments, such a shame
363.
by behnamoh 10mo ago
IQuest stands for it's questionable
364.
by behnamoh 10mo ago
it's not even close to sonnet 4.5, let alone opus.
365.
by behnamoh 10mo ago
Can we donate to creative individuals like the OP so they keep making amazing stuff? This is the kind of output LLMs will not be able to produce any time soon.
366.
by behnamoh 10mo ago
That's not true, the first not-for-humans language is Pel: https://arxiv.org/abs/2505.13453
367.
Gallery of Bad Shell Code(github.com)
3 points by behnamoh 10mo ago|0 comments
368.
by behnamoh 10mo ago
> I feel like we lost humanity somewhere in modern world. I grew up in 3rd world country and if this were to happen, train would literally stop somewhere, anywhere. (With the assumption it’s safe from crewing into another train). While I
369.
by behnamoh 10mo ago
summary: interesting idea, slop website, tested only on old AI models
370.
by behnamoh 10mo ago
Is that the OpenBSD logo they're using?!
371.
by behnamoh 10mo ago
fast but furiously expensive
372.
by behnamoh 10mo ago
I just stopped my Groq API. Sad to see competition being eaten up by shitty Nvidia. I like their products but Jensen is an absolute mfer with deceitful marketing.
373.
Claude Code with API Key?(old.reddit.com)
1 points by behnamoh 10mo ago|0 comments
374.
by behnamoh 10mo ago
Before anyone says "yes you can", read this: https://blog.seemsgood.com/posts/installing-old-versions-wit... In 2025/2026 you can't still simply do `brew install opencode@1.0.190`...
375.
Homebrew still can't install specific version of formula(stackoverflow.com)
5 points by behnamoh 10mo ago|1 comments
376.
by behnamoh 10mo ago
does it violate ISP terms (like at&t)? how to make it less obvious to them?
377.
by behnamoh 10mo ago
I mean, OpenCode has had this feature for a while: https://opencode.ai/docs/lsp/
378.
by behnamoh 10mo ago
No Python LSPs yet!
379.
Nvidia: WTF? (Subscription Computers?) [video](youtube.com)
2 points by behnamoh 10mo ago|3 comments
380.
How to Optimize Your Usage: The Best AI Models to Use(forum.cursor.com)
1 points by behnamoh 10mo ago|0 comments
381.
by behnamoh 10mo ago
> I mean, what's the point of using local models if you can't trust the app itself? and you think ollama doesn't do telemetry/etc. just because it's open source?
382.
by behnamoh 10mo ago
> LMStudio is not open source though, ollama is and why should that affect usage? it's not like ollama users fork the repo before installing it.
383.
by behnamoh 10mo ago
the ability to change email address is not that complicated of a feature to postpone to later. maybe they should ask CC to fix this...
384.
Anthropic: You can't change your Claude account email address(support.claude.com)
89 points by behnamoh 10mo ago|72 comments
385.
by behnamoh 10mo ago
My expectations from M5 Max/Ultra devices: - Something like DGX QSFP link (200Gb/s, 400Gb/s) instead of TB5. Otherwise, the economies of this RDMA setup, while impressive, don't make sense. - Neural accelerators to get p
386.
by behnamoh 10mo ago
I like Anthropic's approach: Haiku, Sonnet, Opus. Haiku is pretty capable still and the name doesn't make me not wanna use it. But Flash is like "Flash Sale". It might still be a great model but my monkey brain associate
387.
by behnamoh 10mo ago
> Don’t let the “flash” name fool you I think it's bad naming on google's part. "flash" implies low quality, fast but not good enough. I get less negative feeling looking at "mini" models.
388.
by behnamoh 10mo ago
> I don't understand how they expected to sustain the advantage against Google's infinite money machine. I ask this question about Nazi Germany. They adopted the Blitkrieg strategy and expanded unsustainably, but it was only a
389.
by behnamoh 10mo ago
> OpenAI made a huge mistake neglecting fast inferencing models. It's a lost battle. It'll always be cheaper to use an open source model hosted by others like together/fireworks/deepinfra/etc. I've been main
390.
Jq Language Description(github.com)
4 points by behnamoh 10mo ago|0 comments

More ›